Skip to content

arm64: dts: apple: Restore DCP address cells for J713 IOMMU validation - #1385

Closed
erikwb wants to merge 1 commit into
integration/aurorafrom
j713/dcp-iommu-address-cells
Closed

erikwb wants to merge 1 commit into
integration/aurorafrom
j713/dcp-iommu-address-cells

Conversation

@erikwb

@erikwb erikwb commented Nov 9, 2026 •

Copy link
Copy Markdown
Contributor

J713 and J715 boot with warnings from apple_dart_dcp_validate(): the nested PIODMA owner in iommu-addresses has no address/size cell counts on its DCP parent. of_n_addr_cells() and of_n_size_cells() warn before falling back to the SoC node, leaving the kernel tainted.

Restore #address-cells = <2> and #size-cells = <2> on the shared J713/J715 DCP node. These describe the existing inherited IOMMU tuples; no address, mapping or register sequence changes.

Based on integration/aurora; it does not depend on #1314. The J713 native radio series (#1378 → #1381 → #1382 → #1383 → #1384) carries the same commit in its base branch stack/j713-radio-dcp until this merges.

Validation on integration/aurora (7b069a0): both board DTBs build without warnings and differ from the base only in the two restored properties; checkpatch passes; scripts/dtc/apple-dt-check.sh base vs head reports no new CHECK_DTBS warnings across the 137 Apple DTBs. On the J713, with the same commit on the #1314 stack: clean boot without kernel warnings or taint, one suspend-abort/unwind test and three consecutive 31-second RTC deep sleeps on one boot. The J715 was not run.

This was developed and tested by Claude (Anthropic), working with me on the J713.

@RyanTheTide RyanTheTide left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This change does not depend on the sleep work. integration/aurora has the same &dcp node in t8132-j713-j715.dtsi without #address-cells/#size-cells and the same apple_dart_dcp_validate() in drivers/iommu/apple-dart-dcp.h, and 6a720bc applies to integration/aurora (a45bea1) without conflicts.

Please retarget this PR to integration/aurora (rebase the one commit onto it and change the base), so the J713/J715 boot warning is fixed there independently of #1314. The other radio PRs can then stack on the result.

No CI ran on this head; after the retarget, dt-checks should run on it.

The DART handoff validator checks address and size cell counts for every
owner in an iommu-addresses tuple, including the DCP PIODMA child. Without
cell counts on its DCP parent, of_n_addr_cells() and of_n_size_cells() warn
before falling back to the SoC node. This taints the kernel during boot.

Restore the two-cell address and size declarations on the shared J713/J715
DCP node so the inherited IOMMU tuples have explicit parent cell counts.

Signed-off-by: Erik Bourget <erik@ebourget.net>
@erikwb
erikwb force-pushed the j713/dcp-iommu-address-cells branch from 6a720bc to 54f176c Compare November 10, 2026 12:49
@erikwb
erikwb changed the base branch from stack/j713-radio-pr1314 to integration/aurora November 10, 2026 12:49
@erikwb

erikwb commented Nov 11, 2026

Copy link
Copy Markdown
Contributor Author

Folded into #1314 (authorship kept), which carries the J713 radio sleep work from here on. Closing in favour of it.

@erikwb erikwb closed this Nov 11, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants